Year: 2020

V7 Labs, the makers of a computer vision platform that helps AI teams “automate” and future-proof their training data workflows as advances in AI continue, has picked up $3 million in funding. Leading the seed round is Amadeus Capital Partners, with participation from Partech, Nathan Benaich’s Air Street Capital and Miele Venture.

Founded in 2018 by Singularity University alumnus Alberto Rizzoli and former R&D lead at RSI, Simon Edwardsson (the same team behind “seeing” app Aipoly), the V7 Labs platform promises to accelerate the creation of high-quality training data by 10-100x. It does this by giving users the ability to build automated image and video data pipelines, organize and version complex data sets, and train and deploy “state-of-the-art” vision AI models.

“For companies to build computer vision solutions that deliver business value, they must continuously collect, label and retrain their models,” explains V7 Labs’ Rizzoli. “When we built Aipoly in 2015, we needed to build and maintain our own tools, whilst keeping up with the rapid state of the art of AI, because no third-party SaaS products were available”.

Fast-forward to today and Rizzoli says that many of the best computer vision companies are now turning to SaaS platforms like V7 to solve this problem. “There’s a lot to think of when building an AI startup, and ‘how can we efficiently store and query 100 different video data sets’ is something you only think of when you’re mid-flight in trying to deliver your service.

“V7 codifies industry best-practices for organizing data, labelling and launching computer vision models for real-world problems”.

The browser and cloud-based platform claims the ability to quickly upload and render large image/video data sets “without lag,” and enable labelling to be automated (to varying degrees) without the need for prior training data. V7 has also been designed to make it possible to keep track of a very large number of labels per image/video, supporting thousands of annotations per image and millions of images per data set. Crucially, Rizzoli tells me it is possible to train, deploy and run computer vision models within the platform “in a few clicks without having to worry about DevOps”.

“Customers will soon be able to audit those models — and their corresponding training sets — to debug, test data quality, discover failure cases and eliminate any unwanted bias,” he adds, noting that these are all huge unsolved pain-points in the AI industry.

To that end, V7 Labs’ existing 100 or so customers include Tractable, GE Healthcare and Merck. It is growing fastest within medical imaging, in part because it offers support for DICOM annotation and HIPAA compliance, both must-haves in healthcare.

However, measured by the quantity of data processed on the platform, Rizzoli tells me that routine “expert inspections” are the most popular tasks. “These include dozens of companies using AI to look for damage or anomalies in cars, oil rigs, power lines, pipelines or roads,” he says.

Chinese tech giants have drawn international criticism after research showed they have technologies that enable the authorities to profile Muslim Uyghurs.

The cloud computing unit of Alibaba, Alibaba Cloud, developed a facial recognition algorithm that can identify a person’s ethnicity or whether a person is “Uyghur”, according to research from surveillance industry publication IPVM.

China has repeatedly defended its controversial “vocational training programs” imposed upon its Muslim ethnic minorities, including Uyghurs, Kazakhs and others, as part of what the government calls counter-terrorism efforts.

Alibaba said in a statement that it is “dismayed” to learn that Alibaba Cloud tested a technology that included “ethnicity as an algorithm” and that “racial or ethnic discrimination or profiling in any form violates Alibaba’s policies and values.”

“We never intended our technology to be used for and will not permit it to be used for targeting specific ethnic groups, and we have eliminated any ethnic tag in our product offering. This trial technology was not deployed by any customer. We do not and will not permit our technology to be used to target or identify specific ethnic groups,” the company added.

A security breach from last year revealed that a “smart city” surveillance system hosted on Alibaba Cloud could detect people’s ethnicity or label them Uyghur Muslim, TechCrunch reported earlier. At the time, Alibaba said as a public cloud provider, it “does not have the right to access the content in the customer database.”

IPVM also found earlier this month that Huawei and artificial intelligence unicorn Megvii, known for its facial recognition product Face++, jointly developed a technology that could alert the Chinese government when the system detected the face of a member from the Uyghur community.

As China’s tech upstarts seek overseas growth, they increasingly find themselves stuck between the demands of Beijing and international scrutiny over their stance on human rights issues.

Cloud computing is one of Alibaba’s fastest-growing segments and the giant is eyeing to attract more international customers. Last year, Alibaba Cloud was the biggest player in the Asia Pacific region and the third-largest Infrastructure as a Service (IaaS) provider globally, according to research firm Gartner.

Alibaba’s cloud unit grew 60% year-over-year to account for nearly 10% of the firm’s revenues in the three months ended September. As of the quarter, approximately 60% of A-share listed companies, those that are based in mainland China and trade in RMB, are customers of Alibaba Cloud, the company claimed.

After a period of public feedback, Twitter adjusted some its plans for a new verification process, set to roll out next year. The company suspended public verification applications in 2017 and since appears to have rethought a few aspects of what information the platform should signal to its users, blue checks and beyond.

One big verification-adjacent change around the corner: Twitter plans to add a way of distinguishing bots and other automated accounts.

“… It can be confusing to people if it’s not clear that these accounts are automated,” the company wrote in a blog post. “In 2021, we’re planning to build a new account type to distinguish automated accounts from human-run accounts to make it easier for people to know what’s a bot and what’s not.”

Of course, not all bots are good bots, but automated accounts have flourished on the platform since its early days and bots remain some of the most useful, whimsical and otherwise beloved sources of tweets.

 

The company is also working on a better way to handle accounts for users who have died, and plans to introduce a memorialization process in 2021. Twitter says that memorialized accounts, like bots, will become “a new account type” making them distinct from normal users. The idea grew out of the same spirit as Twitter’s labels for political figures, which sought to provide contextual info about users that can be seen at a glance.

Taking more than 22,000 pieces of feedback on the new verification process into account, Twitter will no longer require a profile bio or header picture to verify users, calling its former thinking “too restrictive.” It’s also redefined a few of its eligible verification categories, expanding “sports” to include esports and adding more language around digital content creators into the entertainment category.

Twitter also apparently received a lot of suggestions calling for additional verification categories for scientists, academics and religious figures. Until it spins out more categories, those users can seek verification under the “activists, organizers, and other influential individuals” catch-all category.

Verification applicants will need to apply under a particular category and provide links or other information supporting their application. The new “self-serve” verification process will be available through account settings on both mobile and desktop.

Twitter will implement the new account verification policy on January 20, 2021, three years after freezing the process. The company did not specify when public verification applications will be accepted again, but it sounds like the wait won’t be too long and the company plans to share more soon. Starting on the 20th, Twitter will begin sweeping out inactive verified accounts and others that don’t meet its new bar for a “complete account.”

In the adjusted policy, a complete account — and one eligible for verification — must have a verified email or phone number, a profile image and a display name. Anyone who’s verified but doesn’t meet those criteria will receive notifications of the required changes, which must be made before January 20.

Twitter’s new policy also lays out the company’s right to revoke verification for accounts in “severe or repeated violation” of the platform’s rules. It sounds like new policy could lay a clearer path for the company to take against users who break the rules, though that ultimately will come down to enforcement rather than written policies.

“We will continue to evaluate such accounts on a case-by-case basis, and will make improvements in 2021 on the relationship between enforcement of our rules and verification,” Twitter wrote in the post.

Twitter paused the verification process in November, 2017 following a public outcry over its decision to verify Jason Kessler. Kessler infamously organized the Unite the Right event in Charlottesville, Virginia that gathered neo-Nazis and white supremacists, ultimately leaving one peaceful counter protester dead. The pause was extended the next year as the company decided to direct more resources toward election integrity.

With the midterms and the general U.S. election behind it, Twitter has returned to its effort to rethink the verification process and what it symbolizes for users on the platform. The company is also experimenting with new features that could dial down harassment, toxicity and misinformation.

Twitter recently added friction to the retweet process in an effort to slow the spread of misinformation, though it rolled the change back after the election. Twitter’s latest test: A new pop-up that displays shared interests and a profile bio when a user goes to reply to someone they don’t follow.

Autonomous vehicle company Aurora Innovation sent offers Thursday to more than 75% of employees at Uber Advanced Technologies Group, just a week after announcing plans to acquire the self-driving subsidiary, according to a source familiar with the post-merger integration plans.

Uber ATG Toronto, which employs about 50 people where the subsidiary conducted its research and development work, did not not made the cut, according to a source. Nor has Uber ATG’s chief scientist Raquel Urtasun, who led the Uber ATG R&D team. It was previously confirmed that Uber ATG CEO Eric Meyhofer would not join Aurora once the deal closed. Until today, it was unclear if Urtasun, a leading expert in machine perception for self-driving cars and the co-founder of the Vector Institute for AI, would be moving over to Aurora.

Of the 1,200 people who work at Uber ATG, more than 850 received emailed offers from Aurora co-founder and CEO Chris Urmson. In the email, an excerpt of which TechCrunch has viewed, Urmson said the decision of who to pick was difficult. He noted that the decisions were based on Aurora’s specific business needs such as areas of overlaps, relative impact and management reporting.

Aurora wouldn’t comment on the offers, but did confirm that Uber’s Toronto office was not being integrated into the newly combined company. An Uber spokesperson also confirmed that the Toronto R&D lab would not be integrated into the joint organization.

“As an independent company focused on our long-term growth and success, we must be thoughtful about where and how we spend resources. To deliver on our mission, we weave research into our development process and engineering work, rather than having a separate research and development team,” an Aurora spokesperson wrote in an email statement. “We have immense respect for Raquel Urtasun and her team. The impact they have made on both the ATG team, and the industry in general, is incredible. While she and her team will not be continuing on with Aurora, we wish them tremendous success.”

If every Uber ATG employee who received an offer accepts, it would more than double Aurora’s size. Before the acquisition was announced, Aurora had about 600 employees working out of its offices in Palo Alto, San Francisco, Pittsburgh and Texas. Uber ATG had offices in Pittsburgh, San Francisco and Toronto.

Aurora and Uber had been in talks for months before reaching a complex deal that will value the combined company at $10 billion. Aurora is not paying cash for Uber ATG, a company that was valued at $7.25 billion following a $1 billion investment last year from Toyota, DENSO and SoftBank’s Vision Fund. Instead, Uber is handing over its equity in ATG and investing $400 million into Aurora, which will give it a 26% stake in the combined company, according to a filing with the U.S. Securities and Exchange Commission. Shareholders in Uber ATG will now become minority shareholders of Aurora.

At the time the deal was announced, Urmson told TechCrunch that the next 60 days would be spent bringing the two teams together and “dispassionately looking at what is the technology that accelerates our first product to market and then amplifying that.”

Ned Desmond, a longtime publishing executive who spent more than half a dozen years at Time Inc. before becoming the chief operating officer of both TechCrunch and Engadget for more than eight years, has joined the investment firm SOSV as a senior operating partner.

It’s seemingly a good fit for both sides.

SOSV — which is currently managing a $277 million flagship fund alongside some smaller vehicles — has become known for its popular accelerator programs, including Hax, a program focused around nascent hardware startups, and IndieBio, SOSV’s life-sciences-focused accelerator.

In fact, the outfit, founded by serial entrepreneur Sean O’Sullivan in 1995, has now funded so many startups —  roughly 1,000 of them  —  that it recently sought out Desmond to work with them, connect them, shine a light on their work and help them raise follow-on funding.

It’s work that has become intuitive for Desmond, who among other things, was heavily involved in organizing TechCrunch’s multiple events around the world each year, along with its signature Battlefield competitions, which collectively feature dozens of nascent startups annually.

Indeed, soon after he resigned from TechCrunch last summer in search of a new challenge (and some needed downtime), O’Sullivan reached out to him, asking if he might join SOSV to help with its marketing and sales efforts, as well as to provide coaching and connections to its startups.

By O’Sullivan’s telling, SOSV could use the help more than ever.

The outfit has had its share of successes. It was the lead investor in the electric bike company Jump Bikes, acquired by Uber in 2018 for an undisclosed amount. It’s an early investor in the 3D printing company FormLabs (valued at more than $1 billion during its last round in 2018). It also wrote an early check to the peer-to-peer car-sharing company GetAround, which was hard hit by the pandemic but whose business has since reportedly rebounded such that it was able to raise $140 million in Series E funding in October. (It has raised $600 million altogether.)
Not last, SOSV is an investor in the lab-grown meat producer Memphis Meats, which raised $161 million in new funding at the start up of the year led by SoftBank.

Still, SOSV is in the same boat as many seed-stage investment firms. It’s working with a lot of very new teams for whom the pandemic has been rough. According to Crunchbase, seed funding in the third quarter was down 32% percent year over year and down 11% quarter over quarter.

While the buzzier, more established companies have had no problems in fundraising — many are having to bat back overeager investors — newer, unproven teams without established connections have found it harder to land seed-stage and Series A checks.

“It’s a super tough market,” says O’Sullivan. “Angels have completely dropped out. Seed investing is down massively.” Except in life sciences, especially in the U.S., says O’Sullivan, “it’s tougher for every company because you can’t move quite as fast.”

What of Zoom and related talk by VCs of the extra time they now have to meet with new startups? O’Sullivan suggests with a laugh that there’s no shortage of posturing in the industry. “Everyone’s always doing ‘great.’ But I’m a Catholic guy. I have to be honest” and there’s “nothing quite like going to meet with a company and working with them across the table. When you’re working with them remotely, it’s just a slower process.”

With Desmond’s work ethic, his connections and his experience, O’ Sullivan thinks he can speed it up. He’s clear-eyed about the coming months. “The stock market is doing great, but the economy is not doing well, so there’s tremendous uncertainty about what happens next.”

O’Sullivan, is “cautiously optimistic,” he adds, but like a growing number other investors, he’s also wondering, “How long can the party continue?

Google has been all about the fun little augmented reality Easter Eggs lately, with a bevy of search terms triggering Halloween-themed AR experiences back in October. Ghosts! Jack-o-lanterns! Dancing skeletons!

Now they’ve got another one. Rolled out just in time for tomorrow’s Season 2 finale of The Mandalorian, this one brings The Child (or, as the world has taken to calling him, “Baby Yoda”) into your living room.

Want to check it out yourself? You just need to search for the right thing. Here’s how to do it:

1. Open Google.com in the browser of an iOS or Android device
2. Search for “The Child”, “Baby Yoda”, or [The Child’s actual name here which I’ll omit because spoilers]
3. Scroll down until the “View in 3D” button appears. Tap it. (Depending on your device, you might need the Google search app installed. It worked by default on a Pixel.)
4. Wait until the camera view pops up, then wave your camera around a bit. Once the camera figures out where the floor is, he should appear. Be sure to bump your volume up.

I just tested it myself and it worked well, albeit better on some devices than others. It was fast and flawless on a Pixel, but the lighting was bugging out hard on my iPhone.

There’s a camera button that’ll let you snap photos and videos, and you can move him around the room or rotate him in place for better shots. Have fun!

The coronavirus pandemic has underscored, and often exacerbated, the mental health crisis that exists across the world. Even the spread of remote work is part of the problem: As everyone stays at home, the lack of interaction and watercooler chat has left employees without in-person interaction.

The need for a solution has helped tech-powered mental health solutions raise funding to meet increased demand. In the latest development, it emerged that Lyra Health, a platform that focuses on providing workforces with mental health care, has filed paperwork to raise a $175 million Series E at a $2.25 billion valuation.

The paperwork was uncovered by Prime Unicorn Index. While it is not clear whether the company has closed the round, filings in Delaware usually appear after part or all of the funding has been secured. Prime Unicorn Index notes that the terms surrounding this Series E round include a “pari passu liquidation preference with all other preferred, and conventional convertible, meaning they will not participate with common stock if there are remaining proceeds.” It also noted that Lyra Health’s most recent price per share is $27.47, an up round from the Series D, which priced shares at $14.21.

We are reaching out to the company and investors for a response to the filing. One investor noted that the round has not closed yet.

Past backers of the company include Adams Street Partners, Tenaya Capital, Meritech Capital Partners, IVP and Greylock.

We seem to be in a period of rapid growth rounds getting raised in quick succession for the most promising startups. As with Discord — which confirmed a $100 million round just six months after raising $100 million — Lyra Health also recently raised funding — specifically a $110 million Series D that catapulted it above a $1 billion valuation.

That effectively means the startup doubled its valuation in a handful of months, suggesting rapid growth or key validation. As reported by Forbes, Lyra Health was set to bring in around $100 million in revenue by the end of the year at the time of its prior fundraise.

There have been a number of categories of technology that have seen a bump of usage and interest during this coronavirus pandemic, and sadly — or perhaps usefully, depending on how you look at it — mental health and wellness startups, aimed at helping our well-being in this trying time, have been one of them. Just last week, the meditation app Calm raised $75 million at a $2 billion valuation.

Burlingame, California-based Lyra Health wants to live in offices everywhere. The company helps employers give their employees a suite of safe and confidential tools to support their mental health needs. This is a tricky space to play in, considering that mental health can still feel taboo in workplaces and employees might feel uncomfortable turning to their employers for support. Still, in a world where in-office perks are no longer available, mental health might be a key investment to help startup retention.

Once an employee joins Lyra, the company creates a set of recommendations for the now-patient based on a survey. Lyra Health then can connect patients to its network of thousands of therapists for appointments, consultations and check-ins. The flywheel continues.

During the pandemic, Lyra Health has brought on 80,000 new users, to a total of 1.5 million users last reported.

Tech-enabled mental health care has found tailwinds as the coronavirus pandemic leads to a surge of telehealth, as in-person doctor’s appointments could leave patients at risk. Indeed, Lyra Health started Lyra Blended Care, which pairs video therapy with online lessons and exercises rooted in cognitive behavioral therapy.

NASA’s Artemis mission is just starting to get underway, and among the commercial partnerships vying for the privilege of building the lunar landing system is one between Lockheed Martin and Blue Origin, which is leading the effort. Lockheed VP and GM of Commercial Civil Space Lisa Callahan says that the collaboration has been surprisingly smooth and fruitful.

Speaking at TC Sessions: Space, Callahan expressed her excitement for being able to take part in such an endeavor to begin with: “Who wouldn’t want to do that? That’s pretty awesome,” she said. “A lot of our workforce wasn’t around in the Apollo days, so they’re really excited to be a part of this next generation and bringing astronauts back to the moon — and for me personally, the fact that we’re going to bring the first woman to the moon is just amazing.”

She explained that Lockheed is working on the ascent module, while Northrup Grumman and Draper are working on other components, and Blue Origin, the prime contractor, is making the descent module.

“It’s a really fun combination of the entrepreneurial, from the Blue Origin perspective, with some of the heritage companies that Lockheed and Northrop Grumman and Draper provide going back to the Apollo days, to bring a kind of national time together for this national priority,” she said.

One might fairly expect a bit of friction between the old rivals and the newcomer, but according to Callahan it’s been extremely constructive.

“It’s a merging of different cultures, and I think everyone on the team is growing because of it,” she said. “Blue Origin has been a great prime, they’ve really welcomed everybody in a sort of… what I’ll call a badge-less environment. I don’t think if you were sitting in one of the technical interchange meetings that we have, you would even know who works for who. Because we just bring the best of breed and who has the right experiences to do the job we’ve got to the table. So it’s really been quite seamless, and we’ve had a lot of fun with it.”

All despite the pandemic, which has caused nearly every company to change the way it operates. Callahan said that this has really put existing efforts to modernize operations into focus rather than upend their plans.

“We’ve been investing for probably the last five years or more in what we’re calling digital transformation — so, digital collaboration tools, building digital twins of our spacecraft, so multiple people can work on the design at the same time,” she explained. “The silver lining, if you want to think about it that way is… COVID has just helped to accelerate those. It’s teaching us that we can really collaborate in this kind of virtual environment in ways that maybe we’d never thought of.”

Lockheed’s next big milestone is the delivery of its Orion spacecraft to Kennedy Space Station in Cape Canaveral.

“We’re really excited. We’ll be delivering that system over the VAB [NASA’s Vehicle Assembly Building], and it’ll go through its launch prep for a launch that will happen in 2021. And that will be the first time Orion will have launched off of the Space Launch System,” Callahan said.

Missed the event? Extra Crunch subscribers get access to full videos from our stage, including TC Sessions: Space, Disrupt, and all the rest. You can sign up here.

Discord announces a big funding round, Google gets European approval to acquire Fitbit and Twitter launches a new voice-based feature. This is your Daily Crunch for December 17, 2020.

Discord raises $100M

The popular gaming chat platform confirmed today that it has raised $100 million and also announced that it has 140 million monthly active users, twice as many as a year ago.

“We are humbled and honored by the growth we’ve seen among so many incredible and diverse communities that have made Discord their place to hang out,” said CEO Jason Citron in a statement. “As we look to 2021, we are excited about what we have in store and plan to use this funding to help make Discord even better — both for our free service and our Nitro subscribers.”

The confirmation comes after TechCrunch reported that the company was raising up to $140 million at a valuation that could be as high as $7 billion.

The tech giants

Europe clears Google-Fitbit with a ten-year ban on using health data for ads — Under the terms of the EU’s clearance for the deal, Google has committed to not use Fitbit user data in the European Economic Area for ad targeting purposes for a 10-year period.

Twitter launches its voice-based ‘Spaces’ social networking feature into beta testing — During this initial testing period, the product will be limited to select individuals, largely from underrepresented backgrounds, Twitter says.

Google slammed for ‘monopoly power’ in new antitrust lawsuit from 35 states — Compared to the Texas-led suit against Google announced yesterday, the second lawsuit represents a broader coalition of 35 states.

Startups, funding and venture capital

Coinbase files to go public confidentially and we’re hyped — To be clear, I don’t consider myself part of the “we” that’s hyped, but Alex Wilhelm definitely is.

Spryker raises $130M at a $500M+ valuation to provide B2Bs with agile e-commerce tools — Spryker offers a platform to bring a company’s inventory online, as well as tools to analyze and measure how that inventory is selling and where.

Health insurer Oscar adds another $140M in what’s likely a pre-IPO round — The new capital means that Oscar has raised what would be the equivalent of $1 million a day for the entirety of 2020.

Advice and analysis from Extra Crunch

Virgin Orbit, Relativity Space and Astra dish on the economics and efficiencies of space launches — Relativity Space CEO Tim Ellis, Astra CEO Chris Kemp and Virgin Orbit’s VOX Space President Mandy Vaughn all joined us at TC Sessions: Space to discuss their approaches to the small spacecraft launch market.

Just how bad is that hack that hit US government agencies? — Spoiler: It’s a nightmare scenario.

2020’s top 10 enterprise M&A deals totaled a staggering $165B — It was a blockbuster year for enterprise M&A.

(Extra Crunch is our membership program, which aims to democratize information about startups. You can sign up here.)

Everything else

HBO Max finally lands on Roku devices — “Finally” gets overused in headlines, but it absolutely applies here.

You can now securely submit tips to TechCrunch using SecureDrop — We’re making it easier and more secure for you to contact TechCrunch reporters and editors.

The Daily Crunch is TechCrunch’s roundup of our biggest and most important stories. If you’d like to get this delivered to your inbox every day at around 3pm Pacific, you can subscribe here.

UIPath, the robotic process automation startup that has been growing like gangbusters, filed confidential paperwork with the SEC today ahead of a potential IPO.

“UiPath, Inc. today announced that it has submitted a draft registration statement on a confidential basis to the U.S. Securities and Exchange Commission (the “SEC”) for a proposed public offering of its Class A common stock. The number of shares of Class A common stock to be sold and the price range for the proposed offering have not yet been determined. UiPath intends to commence the public offering following completion of the SEC review process, subject to market and other conditions,” the company said in a statement.

The company has raised over $1.2 billion from investors like Accel, CapitalG, Sequoia and others. Its biggest raise was $568 million led by Coatue on an impressive $7 billion valuation in April 2019. It raised another $225 million led by Alkeon Capital last July when its valuation soared to $10.2 billion.

At the time of the July raise, CEO and co-founder Daniel Dines did not shy away from the idea of an IPO, telling me:

“We’re evaluating the market conditions and I wouldn’t say this to be vague, but we haven’t chosen a day that says on this day we’re going public. We’re really in the mindset that says we should be prepared when the market is ready, and I wouldn’t be surprised if that’s in the next 12-18 months,” he said.

This definitely falls within that window. RPA helps companies take highly repetitive manual tasks and automate them. So for example, it could pull a number from an invoice, fill in a number in spreadsheet and send an email to accounts payable, all without a human touching it.

It is a technology that has great appeal right now because it enables companies to take advantage of automation without ripping and replacing their legacy systems. While the company has raised a ton of money, and seen its valuation take off, it will be interesting to see if it will get the same positive reception as companies like Airbnb, C3.ai and Snowflake.